Andy from Carnegie Mellon here. This is impressive work. My PhD student and I have looked into similar methods for autotuning config knobs for MySQL + Postgres:

Andy from Carnegie Mellon here. This is impressive work. My PhD student and I have looked into similar methods for autotuning config knobs for MySQL + Postgres: https://db.cs.cmu.edu/projects/ottertune/ https://db.cs.cmu.edu/projects/ottertune/ One of the many challenges with DB tuning that we have found is that people don't know whether their database, application, or DBMS version has changed enough to warrant another round of tuning. The best values that your tuning produces today may be different two weeks from now. Also, all the knobs that you are tuning don't require restarting the DBMS. Some of the knobs that make the most significant performance difference require you to restart the DBMS first before they take effect, which is not something people want to do often on their production database. None of the knobs that you target require restarting. This problem is super interesting and our research has shown that it can really improve DB performance for some applications. We are in the process of spinning out OtterTune as a new startup: https://ottertune.com/ https://ottertune.com/